Velocity of Sound Waves in Tubes. 213 
tube when 2.4 cm. long, and 0.89 em. when 10 em. long. It would appear 
from this that more than half the gain was made in the first fourth of 
the tube’s length, and that if the tube were long enough the velocity 
might drop to the values obtained by other experimenters, or even below 
—for their results are averages over considerable lengths of tubes. 
The writer gives the calculations above—for Plates I and Il—merely 
as an illustration of what occurred in these two cases, and not because 
he attaches any significance whatever to the numbers given. As a 
matter of fact, the numbers have no significance. In every case I have 
tried, the waves through the tubes have been in advance of those in 
free air, but the gain has been quite variable. I am now endeavoring 
to determine the cause of the increased velocity, and the reasons for 
its variation. I have secured a number of photographs of the waves 
through a 10 em. and a 15 em. tube placed side by side, with their ends 
at different distances from the sound spark. This investigation is not 
complete, but it has gone far enough for me to say that the velocity of 
a pulse through a tube is greatest when the end of the tube is nearest 
the sound spark, indicating that it is a question of sound intensity. 
The sound for a time travels faster in the tube than it does outside 
because the intensity of the wave in the tube decreases less rapidly than 
in free space. 
This experiment appears to settle conclusively the question as to the 
dependence of sound velocity upon intensity independent of any varia- 
tions caused by motion of air in a body, as contended by Rink in the 
case of Regnault’s experiments. I shall discuss in a later paper the 
question of what happens to the air when a spark passes. 
Physics Laboratory, Indiana University, January, 1919. 
